The race-based neighborhood stereotyping hypothesis, proposed by Ingrid Gould Ellen, posits that whites’ residential preferences are driven primarily by race-neutral concerns for neighborhood quality rather than racial prejudice. The theory maintains that whites’ evaluations of neighborhoods may be shaped by stereotypes of the kinds of neighborhoods in which blacks live, but not by stereotypes of black individuals themselves. Using data from the 2004 Detroit Area Study, this paper tests the race-based neighborhood stereotyping hypothesis as an explanation for whites’ residential preferences. In support for the theory, results show that neighborhood stereotypes mediate the effect of racial composition on whites’ ratings of neighborhood desirability. The extent to which these neighborhood stereotypes are distinct from stereotypes towards individuals, however, is questioned. Results indicate that, rather than representing a shift away from negative affect towards blacks themselves, neighborhood stereotypes may simply be another method of expressing continued racial prejudice.